gpt-6POST
Will Depue speculates OpenAI's new model is a fresh pretrain, not RL fine-tuning
He judges from repeated hints by OpenAI staff that it is a new pretrained model, saying it surpassed Astra and solved Navier-Stokes less than two weeks into pretraining.
will depue
@willdepueArchived 2026.09.16Originally posted 2026.09.15
will depue @willdepueOriginally posted 2026.09.15
openai people keep gesticulating towards the ‘new model‘ so much so that i’m absolutely sure it’s a new pretrain and not some RL run, which implies (i think) OpenAI has a model that not more than 2 weeks into pretraining was better than Astra and solved Navier-Stokes. insane
This post quotes another post: x.com/polynoamial/status/2099726370356314563
x.com/willdepue/status/2099746385486635469Archived 2026.09.16 10:17